Coffee and Psychopaths and Clickbait

Today I’m going to get into the article behind a story that has been doing the rounds on Facebook recently. The headline claims that Science (with a capital S) has discovered how psychopaths drink their coffee, the accompanying picture is a strong, black cup of coffee. It’s a great click bait headline inspiring mock righteousness and indignation from all areas of the coffee spectrum. While this seems a frivolous topic (and the story I read certainly made it out to appear that way) the original study is situated in a broader area of personality research which seeks to find connections between taste preferences and individual personality traits.

Individual differences in bitter taste preferences are associated with antisocial personality traits by Christina Sagioglou and Tobias Greitemeyer from the University of Innsbruck in Austria.

Taste preferences are mainly believed to serve evolutionary needs (sweet foods often have the high calories needed for survival, while bitter foods often indicate toxic substances). But beyond these most basic needs, the food we are drawn to is influenced by many other things. The first taste of coffee or beer, for many people, is disgusting, but over time, people can grow to enjoy it. But this involves a period of adjustment, where one must forgo one motivation (that of pleasant taste) to experience another (social, neurochemical, or health related). Sagioglou and Greitemeyer wondered if it was a certain type of personality that persisted with bitter foods in order to gain a taste for them.

Several research studies suggest that personality traits can predict what food people choose to eat. People who have a high sensation seeking personality (think people who are always busy and trying new things) tend to prefer spicy food and foods high in caffeine more than those who don’t have that personality. Those who are very agreeable tend to like sweet foods compared to those who are less agreeable. Researchers have even been able to predict whether someone would select a glass of dry white wine over a sweet white wine based on how open to experiences they are. So there is definitely enough research to suggest personality and taste preference may go together.

This paper also raises an interesting possibility. They draw on past research that suggests, not only can personality traits influence taste preference, but continued exposure to particular tastes may increase corresponding personality traits. If a person was given something sweet they claimed to feel more agreeable. By contrast, when people were given bitter things to eat, they became hostile and argumentative! Sagioglou and Greitemeyer propose that prolonged exposure to bitter tastes could extend hostility into a personality trait.

Because of this background of research Sagioglou and Greitemeyer sought to determine if a preference for bitter tastes was associated with antisocial personality traits. They conducted two separate studies to explore their idea.

The Research

This paper consisted of two studies that explored the same question. Both were conducted on Amazon Mechanical Turk (MTurk) with sample sizes of 449 (214 female, 235 male) and 504 (247 female, 257 male) respectively. In both studies, participants were presented with a list of sweet, bitter, sour, and salty foods (included at the bottom of this entry) and were asked to rate how much they liked each food (from “dislike strongly” to “like strongly”). They were also asked how much they liked sweet, bitter, sour, and salty tastes in general.

In the second study participants were also asked to give a taste profile to each food presented. This was done to check whether the food categorised as bitter was actually perceived as bitter. Because a black shot of espresso is very different in taste to a Starbucks caramel latte, and they wanted to make sure they accounted for the differences in personal ratings. Indeed, when the researchers went back and looked at the foods they had labelled at ‘bitter’ only five of the original ten were rated as bitter foods and included in the analysis. These included: coffee, beer, radishes, tonic water, and celery.

For personality measures, participants gave details on their levels of verbal and physical aggression, anger, and hostility. They completed a measure of dark triad personality traits (Machiavelism, psychopathy, and narcissism) as well as a measure of sadistic tendencies. Finally, they completed a short personality measure which looked at the Big Five* personality traits (openness, conscientiousness, extraversion, agreeableness, and neuroticism). These measures were the same in both studies.


As predicted, the people who said they liked the bitter foods, as opposed to those who did not like them, were more likely to have higher scores of psychopathy, narcissism, everyday sadism, and aggression. They were also more likely to score low on the measure of agreeableness. These results were similar in the second study, with a couple of differences. The relationship between liking bitter foods and aggression disappeared, and one between bitter taste preferences and Machiavellianism became statistically significant.

When looking at particular antisocial personality traits, the researchers did find a small association between preferences for bitter foods, psychopathy, and sadism. And when I say small, I mean small. The statistics presented reveal that, in their sample, liking bitter food accounted for just over 1% of the variability in psychopathy and sadism scores in the first study, and psychopathy, sadism, narcissism, and Machiavellianism in the second study.


The difference in results between study 1 and study 2 is likely due to the re-evaluation of what actually was perceived as a bitter food by participants. While this adjustment didn’t make the relationship any stronger, it may have led to slightly cleaner results. If you are going to put stock in any of the relationships I would pay attention to the ones that were present in both study 1 and study 2.

What does all this mean?

This study isn’t perfect, and the researchers admit that. One thing they bring up is that using self-report to measure people’s taste preference has some flaws. People prepare foods in different ways, which affects the taste they personally experience. The researchers took this into consideration in the second study by collecting information on the personal taste profile people assigned to different foods and this did lead to eliminating half of what they originally categorised as bitter. For example, in the first study tea was grouped with the bitter foods, but when the researchers looked at what people actually thought tea tasted like, it did not receive a bitter rating. Many people add milk, sugar, or honey to their tea which drastically changes their experience of taste. One way to address this is to conduct a more rigorous controlled experiment. Giving participants a dropper of clear solution removes a host of other factors that can impact a person’s perception of taste**. However, I won’t criticise the self-report aspect of this study too harshly.

Another problem with self-report research is the social desirability bias. Even though these surveys are carried out anonymously there is a psychological cost to answering personal questions. As social creatures, humans are motivated to be liked by other people. In this particular study, this could dampen reporting on antisocial personality traits. The researchers also speculated if the linguistic association between the word “bitter” and negative personality traits may have also impacted likeability ratings of bitter foods.

However, I won’t criticise the self-report aspect of this study too harshly. The relationship between taste preference and personality is a relatively small body of research. While highly controlled experiments are the gold standard for understanding causal relationships, they are also expensive, time-consuming, and can easily be a huge waste of time if you don’t have a solid foundation of research to build upon. Research like the current study is important to test if a relationship is even worth exploring further. While the link between bitter taste preferences and anti-social behaviours doesn’t appear to be very strong the results suggest delving further into this link may be fruitful.

Back to the media

Now we come back to the scores of articles that have been claiming that “Psychopaths like to drink black coffee, “How you drink your coffee ‘could point to psychopathic tendencies“, “People Who Order Coffee Black Are More Likely To Be Psychopaths“. One thing I couldn’t help thinking while I was reading the original research this latest click bait craze was based on was, “I wonder why no one is claiming liking radishes makes you a psychopath?”. Of course, the answer to this is that no one would click on that headline. I did see some that latched onto IPA’s and gin and tonic.

The buzz around this study (which is over 2 years old by the way) is exactly why I started this blog. The original research barely speaks of coffee, let alone it’s predictive value when it comes to psychopaths. It takes a very creative journalist to read the research I did and come to the conclusion that all those articles did. Even without an understanding of scientific methodology and statistical analysis Sagioglou and Greitemeyer presented a measured conclusion about their findings.

What you need to take away from this: Drinking black coffee does not make you a psychopath. The most gracious interpretation of these findings is that a preference for bitter tastes may be one of the many factors that are different between people who have more anti-social personality traits than others.

If you enjoy your coffee black, your chocolate dark, and your drinks heavy with the tonic continue as usual safe in the knowledge that you are not devoid of human emotion. However, if you prefer IPA’s I will continue to judge you.

*I talked about the Big Five personality traits in Are Grammar Police really that bad? and I’ve included a link below if you want to learn more about them.

**Research into the psychology of taste perception is amazing! You may think that it’s your taste buds that are responsible for how you experience flavour but there’s so much more. The sound of crunching can make things taste fresher. The colour of a liquid can affect the flavour of a drink. Even the words used to describe a food can make it taste it different. Let me know if you would like to hear more about this kind of research and I’ll add it to my list of future Research Ship destinations. 


Original Article

Individual differences in bitter taste preferences are associated with antisocial personality traits (Christina Sagioglou and Tobias Greitemeyer)

I apologise that today’s article is behind a paywall. I always do a thorough search to try and locate science that the general public can access because science should be free for everyone. I’ve attempted to give as much information as possible about the study, but if you have any further questions I would be happy to answer them in the comments.

List of Foods Used in the Study

Sweet: Candy; Caramel; Chocolate cake; Honey; Ice cream; Maple syrup; Pears; Raisins; Strawberries; Sugar.

Bitter: Beer; Celery; Coffee; Cottage cheese; Ginger ale; Grapefruit; Radishes; Rye bread; Tea; Tonic water.

Sour: Cranberries; Granny Smith apples; Lemons; Lemon drops; Limes: Lime sherbet; Plain yoghurt; Sauerkraut; Sour cream; Vinegar.

Salty: Bacon; Beef jerky; Caviar; Dill pickles; Green olives; Pretzels; Salt; Saltine crackers; Salty peanuts; Soy sauce.


The Toxicity of the American Dream

Today’s post is going to be something different. I wrote this essay for a class I took last quarter that explored the psychology of poverty and income inequality. I wanted to share it with my readers because I think it touches on things that we think are true, but don’t have the resources to back up our arguments. I have adjusted it slightly to make it easier to read (APA format is not the nicest format to read), and arranged the sources I used into categories to make it easier to track down the research if you are interested. I’m excited to hear what you think.

The road to success is not easy to navigate, but with hard work, drive, and passion, it’s possible to achieve the American Dream.

Tommy Hilfiger, Fashion Mogul

It is a fundamental idea of the American Dream that with enough work and dedication anyone can succeed. That everyone can ascend the class ladder by rolling up their sleeves, putting their nose to the grindstone, and pulling themselves up by their bootstraps. With such an optimistic narrative, we should expect that working-class Americans, with ambitions of higher social status, would be buoyed by this national folk-tale. However, low-income individuals routinely show lower levels of psychological well-being and self-esteem. Psychological research into attributions of poverty, system justification, and just world beliefs suggest that it is precisely the American Dream that is leading to low-income people’s low self-esteem. In this essay, I will explore the reasons why the American Dream is so engaging, especially to those of low social status, and the psychological consequences of endorsing this ideal.

Belief in the American Dream

The American Dream is a nebulous concept that includes the ideals of hard work, talent, equality, loyalty, opportunity, wealth, power, and homeownership. There is much overlap between the American Dream and the notion of meritocracy. The foundation of meritocracy is that opportunities are available equally to all members of a society regardless of status, race, or gender. And that every person will gain equitable outcomes based on the effort they put in. The 2014 General Social Survey found 69% of Americans surveyed believed that it is through hard work that people get ahead in life. Belief and endorsement of the American meritocracy have been widely described in the social psychology literature. Researchers, McCoy and Major, found that both men and women, from low and high-status groups could be primed with meritocratic beliefs which affected attributions of unfair behaviour. Others found that belief in the meritocracy predicts a number of gender hierarchy-legitimizing beliefs found in both men and women. This body of research shows support by both high and low-status individuals supporting meritocratic beliefs in both student and community samples.

Despite its widespread belief, the American Dream is far from accurate. The reality of everyday American’s being successful through hard work and dedication may as well be a dream. Indeed, the U.S. has one of the highest levels of income inequality in the world and is among the lowest for social mobility. In terms of educational attainment, though students from all social classes, genders, and races should be given equal opportunity to succeed in school, this is not the case. Schools in different districts and students within the same schools are treated different depending on external factors regardless of how much talent and effort they show. There remains inequality in wages and career attainment across genders and races. The top 1% of Americans earns over 184 times the bottom 90% in total. So what is it that keeps the American Dream alive in the face of this disturbing reality?

Beliefs About Poverty

There are different ways people can view poverty. These can be due to individualistic or structuralist reasons. Individualistic reasons attribute poverty to internal characteristics such as laziness and stupidity. By contrast, structuralist reasons use external attributions such as luck and systematic prejudice to explain poverty. These views have real-world impacts on prejudice and support for poverty ameliorating policy. Researchers, Bullock, Williams, and Limbert, found that people who believed poverty was due to structuralist factors (i.e. that everyone is not given an equal chance, the rich take advantage of the poor, and there is discrimination in hiring and promotion) were more likely to believe that wealth comes from privilege, they were more likely to dislike inequality, and were more supportive of progressive wealth redistribution policies. Those people who believed poverty was due to individualistic factors (i.e that the poor are lazy and less intelligent, and people are poor because they are picky and don’t take difficult jobs) were more likely to believe that wealth comes from ambition and perseverance and were more supportive of restrictive punitive policies towards poverty (e.g. fingerprinting welfare recipients).

For most lower-income Americans endorsing systematic explanations for income inequality is self-protective. When thinking about why you are poor you can choose from either individualistic explanations, which blames yourself, or you can use structuralist explanations, which places the blame elsewhere. This is analogous to making internal versus external attributions. There is a large body of research which shows that making internal attributions about negative characteristics can lead to lowered self-esteem and increased depressive symptoms. As such, it is unsurprising that low-income individuals would seek to make external (or structuralist) attributions for their low-class status.

A study by researchers Major, Kaiser, and McCoy in 2003 illustrates the protective nature of structuralist attribution. When students were told they were not accepted into a class because the professor thought they were stupid (i.e. individualistic attribution) as opposed to the professor being sexist (i.e. structuralist attribution), they were more likely to make internal attributions and blame themselves. They were also more likely to have depressed and anxious emotions. In fact, the less people accounted for their rejection in terms of discrimination, the more they blamed themselves, and the more depressed they were. This suggests that when people can explain a personal rejection because of prejudice (or in the case of poverty a broken system), they are able to make external attributions which discounts personal responsibility. This, in turn, leads to less negative psychological effects.

System Justification and Just World Beliefs

The bulk of research has found that high-status groups overwhelmingly support individualistic explanations, preferring to blame poor people for their situation, while low-status individuals are more likely to support structuralist views. If this were a universal truth only rich people would believe in the American Dream. But we find that low-status individuals also hold individualistic views about poverty. If upholding the dominant ideology of the American Dream is detrimental to low-income people, why would they still hold these beliefs? Research into just world beliefs and system justification has given some suggestions as to why this may be occurring.

Just World Beliefs

Individuals are motivated to maintain their belief in a just and fair world. When threats to this worldview arise (e.g., working hard for no results) we are motivated to shift blame away from the system and as such continue to believe the world is fair and that people get what they deserve. As such, people are more likely to blame themselves for perceived failure if they have high belief in a just world than if these beliefs are low. When an individual looks inwards for the reason why a negative outcome has occurred they are more likely to have negative psychological outcomes such as lowered self-esteem.

Group identification has also been shown to play a role in the effects of just world beliefs. The extent to which someone identifies with their minority group impacts how just world beliefs affect their self-esteem. Research led by Miguel Ramos in 2014 found that after hearing about an innocent victim that shared the same group identity as themselves those who had high just world beliefs scored lower on a state self-esteem measure than those who had a different group identity. That group identification matters is an important distinction when discussing just world beliefs and poverty. Most research into just world beliefs and identity is conducted using race and gender. While stigma surrounds both of these identities, their salience often fluctuates depending on the situation. When we explore how just world beliefs impact those in poverty it is important to be aware of how the stigma of poverty is salient almost constantly. There is a large amount of stigma surrounding poverty, which is compounded when people believe the reasons for poverty are self-imposed. As an illustration, a field study conducted by Karen Seccombe, Delores James and Kimberly Battle looked at the attitudes of single mothers currently receiving welfare. They found the majority of those interviewed were aware of the stigma surrounding their welfare status and routinely experienced the negative consequences of this stigma in their daily lives. This suggests that identification with the low-income group is often salient to those who experience poverty and difficult to de-identify with. Given that the more someone holds just world beliefs and is highly identified with their minority group the lower their self-esteem, understanding the effect of group identity on low-income individuals is an important part of interpreting just world beliefs.

System Justification

Another explanation for why low-income people might endorse the meritocracy is a concept known as system justification. System justification suggests that people are motivated to justify the status quo in order to reduce threat. The more people are likely to crave a structured view of the world, see the world in terms of black and white, and have a fear of death, the more they are likely to uphold the current system. There is evidence to suggest that endorsing the current system can be self-protective in that it increases satisfaction with your current position in life and reduces feelings of guilt and frustration.

But there is also evidence that people who do not benefit from the current system (such as poor African Americans) and yet still hold system justification beliefs can suffer from negative psychological outcomes. By endorsing the current system, disadvantaged groups are more likely to show increased out-group favouritism. This means that they will display greater favouritism towards those who are already privileged by the system despite not belonging to that group. For example, African Americans show greater favouritism towards White Americans than to their own group.

Effects on Self-Esteem

Throughout this large body of literature there are differing views on whether holding just world beliefs and system justification is detrimental or has self-protective effects. The next section will explore both the good and bad of endorsing the American Dream.

The Good

Research led by Shannon McCoy in 2013 found that belief in the meritocracy had an overall positive effect on self-esteem. She and her collaborators reasoned that meritocratic beliefs serve a self-protective purpose, allowing people of low status to regain a sense of control over their future — they may not have anything now, but if they work hard enough, they can look forward to a better life. Further, they explored the different reasons why low and high-status individuals practice system justification. Wealthy individuals are motivated to endorse the meritocracy because it justifies their high status. This is an ego protective measure which says they are where they are in society because they are better than other people. For low-status individuals, however, the perception of control affected the relationship between meritocracy beliefs and self-esteem. Those who felt they had a lot of control over their life had boosts to their self-esteem if they believed in the meritocracy. Whereas those who felt they did not have control did not see this positive effect. Other research has also found that when students received a lower grade than expected those that had higher levels of just world beliefs made less external attributions (i.e. they were more likely to blame themselves). This led them to perceive the system as less unfair resulting in more positive emotions as they felt that they had more control over the situation (i.e. they could study harder next time).

Another line of research suggests that system justification is a means of reducing existential threat and uncertainty. People are motivated to believe in the idea of a meritocratic society because it helps to justify the inequalities that are present in our society. This hypothesis states that people are more likely to engage in system justification practices when the system is questioned. Research conducted by Alison Ledgerwood and her colleagues found that individuals were more likely to judge an article supporting meritocracy as having more sound evidence than an article that challenged meritocracy. Furthermore, this effect was strongest in those participants who had read the anti-meritocracy article first and as such were feeling more threat to the system. Given that existential threat has been shown to have negative psychological consequences this line of research suggests that upholding belief in the American Dream can be self-protective.

The Bad

Not all research into the psychological effects surrounding attributions of poverty has found these self-protective effects. A field study of 100 African American single mothers on welfare found that over half of those interviewed believed in the American Dream. These respondents believed it was their fault they were on welfare and admitted “they had not looked very hard for a job”. These participants also reported feeling more stigmatised and were less assertive in relation to their rights as welfare recipients. These beliefs were also related to lower levels of self-esteem and higher levels of depression and self-harm. The women who held closely to the belief in the American Dream had significantly lower levels of self-esteem than those who believed their poverty was due to systematic reasons.

Researchers, O’Brien and Major, found mixed results for system justification amongst low-status minority groups. Their research focused on the level of identification African Americans and Latinos had to their minority identity and system justification beliefs on self-esteem. They found that for minorities who did not identify strongly with their minority identity, system justification beliefs had the same effect on self-esteem as it did for high-status Caucasian’s (i.e. it was related to higher levels of self-esteem). However, for those who were strongly identified as their minority identity holding system-justifying beliefs was related to lower levels of self-esteem.

A study into women’s belief of the protestant work ethic (working hard for the desired outcome) found that women who perceived themselves to be overweight and held high beliefs about the protestant work ethic had lower self-esteem than those who did not endorse the protestant work ethic. Conversely, those who perceived their weight to be normal and had higher belief in the protestant work ethic had higher levels of self-esteem. While this article explored stigma associated with being overweight, it is a relevant comparison given that both poverty and weight are viewed in society as being under the control of the poor/overweight individual. This maps on to what we know about socioeconomic status and system justification. Those who are successful in terms of socioeconomic status and hold positive beliefs about the meritocracy have higher self-esteem, but not those who are in the lower socioeconomic brackets. Women who believed they were overweight and that they had control over their weight had lower levels of self-esteem than those who did not believe they had control. As with beliefs in the protestant work ethic, the more control women felt they had, the higher their levels of self-esteem.

More research has found that not only were people’s self-esteem negatively affected by negative situations they had no control over, these negative events led to them believing they deserved those events and that they deserve more negative events to befall them. Further, this led to self-handicapping in future behaviours, thus creating worse outcomes. In the case of people in poverty, this creates a self-perpetuating cycle of system justification that feeds further low self-esteem. One negative event (e.g. losing a job due to a company downsize) lowers self-esteem, this is rationalised as being deserved, thus justifying the system, which in turn keep self-esteem low and leads to self-destructive behaviours such as self-handicapping (e.g. not applying for a job they think they won’t get). Not only does this continue to cycle of negative psychological outcomes in low-income people, but it creates another obstacle on the path to economic security.

Furthermore, researchers, Foster, Sloto, and Ruby, tested the direct correlation between belief in the meritocracy and self-esteem. The found that for participants who had experienced discrimination in the past the more they endorsed the idea of the meritocracy the lower their self-esteem. This research was replicated in Latino students which found both correlational and causal evidence that after being exposed to discrimination, Latino students who held high levels of meritocracy beliefs had lower levels of personal self-esteem, compared to those who did not endorse the meritocracy. Further to this, students high in meritocracy beliefs were more likely to blame their own Latino community for their low status in America.


While some research has supported the notion that supporting the idea of the American Dream can be self-protective there is equally as much research which shows the negative impacts of low-status groups holding such beliefs. The evidence supporting the protective effects of just world beliefs and system justification appears to be driven by population groups containing large amounts of high-status individuals. It does appear to the be case that people from high-status groups (i.e. wealthy, white, and male) do indeed gain a threat-protective function from system justification and holding just world beliefs. However, when the data is broken down into groups containing high and low-status individuals, the effects look starkly different. Overwhelming, research has shown that those in lower status positions experience lower self-esteem when they endorse beliefs akin to the American Dream — that hard work and dedication will result in success. These different findings have been attributed to varying levels of perceptions of control and minority group identification. There are likely other factors that account for why some people gain protective effects from system justification while others do not, and further research should address these gaps in knowledge. What is clear is that while the rhetoric of the American Dream, that hard work and perseverance leads to economic gains, continues to hold a place in the forefront of American culture it is critical to examine the psychological impact this is having on our most vulnerable citizens.


Exploring the Meritocracy & the American Dream

Facts on Income Inequality & Poverty

Psychology of Poverty

More on System Justification & Just World Beliefs

Negative effects of poverty on self-esteem

Meritocratic Beliefs & Self Esteem

The Good

The Bad

I Confess! … (even though I’m innocent)

If you follow court proceedings, you will be aware of how important confessions of guilt are in securing a conviction. In Netflix’s hugely popular Making of a Murderer we saw first hand how, once a confession is obtained, there is very little that can be done to move the investigation and conviction in another direction. In the cases where the defendant is indeed guilty of the crime, this isn’t a problem. But it is shockingly easy for interrogation techniques to lead to false confessions, which then lead to false imprisonments. And when you get an innocent person locked away, you also get a guilty criminal free to commit more crime. Today I read a Wired article that discussed interrogation techniques and it started me down a rabbit hole of false memory research. Today I’m presenting one of those research papers:

Constructing rich false memories of committing crime by Julie Shaw (University of Bedfordshire) and Stephen Porter (University of British Columbia).

Shaw and Porter were aware of a large body of anecdotal and case study evidence from real-life police investigations and criminal trials that showed how false memories could occur from police interrogation. While these stories are important at substantiating the effect in the real world, it is just as important to test empirically if it is possible to plant a false memory which, in turn, leads to a false confession.

The Study

Sixty university students and their parents participated in this study (the study originally had ninety, but the effect was so strong the study was stopped early). The inclusion of their parents allowed the researchers to get information about the student’s childhood including a highly emotional event that had occurred during their child’s early teens. It also allowed researchers to leverage this parental relationship when attempting to implant the false memory.

Once Shaw and Porter had the details of this emotional event, the brought the students back into the lab for three interviews over three weeks (one per week). During these interviews, the researchers asked the students about two emotional events that had occurred during their childhood. One of the events was true, while the other was false. The inclusion of a true memory was to confirm the cover story used that the research was into memory recall, rather than false memory implantation. Some examples of the false events they questioned about include, assaulting another child with a weapon, being attacked by a dog, stealing something and been questioned by the police, and having a severe accident.

Unsurprisingly, during the first interview, the students, while remembering the true highly emotional event, didn’t recall the false event ever happening. To ensure the false event was implanted in memory the researcher provided cues that were supplied by the students’ parents such as the time of year and how old they were when the event occurred. To further prompt false memories to form, the researchers told the students that “most people can remember these kinds of memories if they try hard enough”. They also told the students to try and visualise the false event. 

If this seems slightly creepy and coercive, that’s because it is. But that is exactly what happens during police interrogations.

The Reid method of interrogation, which is widely used throughout US police forces, instructs police to create a “theme” to the interrogation. This boils down to creating a story about how the suspect is guilty of the crime and continuing to present this story to the suspect regardless of what the suspect says. Shaw and Porter developed their interview style to mimic the way police interrogate suspects as they were aiming to test if false memories and false confessions could be created in this way. Some other methods they borrowed from police interrogations include presenting incontrovertible false evidence (“your parents said…”), suggesting that they had more information but not being able to reveal it to the student (“this sounds like what your parents described”), and looking disappointed when the student couldn’t recall the false memory.

When students came back for their follow-up interviews two and three weeks later, the same interviews were conducted again – with the researchers prompting the students to recall as much as they could about the true and the false memory, using the techniques outlined above. Remember, this study lasted three weeks. The students were questioned about these false memories three times.

What was the outcome?

Seventy percent of the students who were told they had assaulted another child or stolen something believed this memory after the three interviews. Not only did they believe it happened, a majority of them were able to describe the (fictional) police that they met during that event without prompting. When students were given a false memory that didn’t relate to police contact (e.g., being attacked by a dog, losing money, or having an accident) the percentage of students who believed the false memory was slightly higher at 76%, but this difference isn’t statistically different to the 70% who believed crime related false memories.

These results blew my mind. When you think about the pressure that is on suspects during a criminal investigation, you can make justifications as to how a few vulnerable people might give in to the story that is being constantly repeated at them. But to take a group of university students, with no criminal background, and conduct interviews with them in the guise of a study on memory recall and get three-quarters of them internalise a memory that is traumatic and in some cases, casts them in a negative light is truly shocking.

This study also serves to show how important it is to take an effect that is well known in the field and test it in an empirical setting.

Some good news to end on. The Wired article that got me thinking about false confessions goes into detail about a new wave of interrogation that is backed by empirical research. It is still early in its implementation. But as a new generation of police interviewers comes up through the ranks, the good old days of coerced false confessions may be coming to an end.


The main paper discussed

Constructing rich false memories of committing crime (Julie Shaw and Stephen Porter) 
Version not behind a paywall

The article that started me down this train of thought today

A Severed Head, Two Cops, and the Radical Future of Interrogation


Check Your Privilege!

I spend a lot of time on the internet. During this time I often come across debates about social injustices that very much do need to be addressed. One phrase I’m sure anyone who is familiar with these parts of the internet will be aware of is “check your privilege”. Sometimes it is used as a dramatic line before the commenter slams the metaphorical door in someones face. Other times it is more well intentioned. When used in this way, I’m assuming, the intention is to make clear that other people may be coming from a place of privilege that others do not have, be that because of race, gender, sexuality, economic background etc. Today’s article isn’t going to focus on what privilege is, or how it impacts social interactions. But rather, I’m going to explore a piece of research that popped up in an Intergroup Relationships lecture I took a few years back. It stays forefront in my mind, especially when I see someone imploring another to “Check their privilege”.

Racial Attitudes in response to thoughts of white privilege by Nyla Branscombe, Michael Schmitt, and Kristin Schiffhauer from the University of Kansas in the USA and Simon Fraser University in Canada.

This particular study looked at how white Americans would act when they were confronted by their privileged position within society.

The authors suspected that even though it is well known that white people in America overall do better economically and socially, most white Americans do not see this as privilege. So when they are told to “check their privilege”, this can lead to a sense of imbalance in their views of themselves. Most people see themselves as being good and moral but when made to think of their privilege they must also comes to terms with having what they have because another group does not. This creates a very threatening state of mind. The authors put forward that people might respond to this threatening state by being MORE racist. Which is entirely the opposite of what those who chant “check your privilege” want.

Their predictions were based on a theory known in social psychology as Social Identity Theory. This theory basically says that the social groups we belong to are important parts of our personal identity. So we become intertwined with how those social groups are depicted and it can directly affect our self esteem. The argument this paper is making is that by showing people that the only reason their social group is successful is because it oppresses another group, this threatens not only the social standing of the group in question, but of all the individuals that are part of that group.

There are many ways in which individuals can deal with this threat. Ideally, they will change their behaviour and own the moral decision to do so. But this requires a lot of effort and brains are typically pretty lazy (stay tuned for research into that giant subject). What Branscombe and her colleagues argued was that people would find a way to legitimise the privilege their group has. They would argue that they weren’t privileged because of the oppression of another group, but because they worked harder and were more deserving. And that this would ultimately lead to more racist behaviours.

Modern Racism compared with Explicit Racism.

Before I get too deep into what actually went on during this study, I’m going to take a moment to describe different forms of racism. I don’t think there is any doubt that racism exists. What I don’t think is as clear is that racism is not always as overt as apartheid regimes and offensive media stereotypes. These examples, and probably the things you think of when you hear the term ‘racism’ are types of Explicit Racism. These are the obvious outward prejudices people display to anyone of another race. In some senses, explicit racism is easier to deal with. It stands out clearly and can be combatted with institutional initiatives such as Affirmative Action.

Modern Racism, on the other hand, is much more insidious. Also called Implicit Racism, this type of racism manifests itself as a belief that racism is a problem of the past. That we have equality today, and that any differences between the races are a matter of biology not institutionalised racism. You can see that this is harder to tackle. It creates its own circular argument. When you try to point out that this is a type of racism – the person displaying modern racism will retaliate that racism no longer exists. This is the racism that leads to backlash against #blacklivesmatter, that supports removing all affirmative action, and overall ignores insitutionalised racism despite the swathes of research that backs up its presence. 

And it is modern racism that the mantra “check your privilege” is often levelled at. Which is why this study is so important.

Back to the study

The paper covered two separate studies that together paint a very interesting picture about how thinking about one’s privilege affects modern racism. In the first study, participants were asked to either:

  1. Write about the ways they benefited by being white
  2. Write about the ways they were disadvantaged by being white
  3. Write about general life experiences.

After participants had written as much as they could they answered a series of questions designed to measure their levels of modern racism. Some of the items in this scale include questions such as:

  • Discrimination against Blacks in no longer a problem in the United States
  • Blacks are getting too demanding in their push for equal rights
  • Over the past few years Blacks have gotten more economically than they deserve

When the experimenter’s looked at what had been written initially they noticed that overall participants were able to come up with more ways in which they were privileged than disadvantaged. Something that struck me as interesting when reading this article was that even though they weren’t instructed to think about their privilege over black Americans specifically, the majority of participants brought this up in their lists. For those thinking of how they were privileged they “almost exclusively focused on avoiding the disadvantages of being Black.” This isn’t something that the researcher’s were expecting, or were able to analyse statistically, but it is an interesting outcome to think about.

What was expected by the study was that when these white participants spent time thinking about the ways they benefited from their privileged status they showed statistically more modern racism than those who thought about disadvantages or who simply wrote about their life. Furthermore, because there was a difference in the number of thoughts generated in the privilege compared to the disadvantage task, the researchers ran further statistical analyses and concluded that this did not explain the greater level of modern racism seen in those who focused on their privilege.

This first experiment alone was enough to make me never want to tell someone to “check their privilege” ever again, but the second study shows something even more fascinating. This is where the researchers get into issues of individual differences in people.

Branscombe and her colleagues thought that if being made aware of your privilege as a white person is threatening your identity and therefore makes you display more modern racism, it should follow that those who identify strongly as white will be more likely to feel that threat, and as such would display more racism. That is what they set out to explore in the second study.

Similar to the first study, they had participants conduct the same thought tasks (write about your privilege, write about your disadvantages, or write about your general life experiences). They also had participants answer some questions about their racial and political identification.

Because this study used the same methods as the first one, the researchers were able strengthen their findings by replicating their results. Again, overall, those participants who were asked to think about how white privilege affected them, showed statistically higher levels of modern racism than those who did not.

On top of this, those participants who strongly identified as white, and spent time thinking about their privilege showed higher levels of modern racism still. The story is more complicated than this. What is actually going on is that thinking about privilege affects those who have high or low white identification differently. Those who highly identify as white react strongly when made to think about privilege – these are the people who show higher levels of modern racism. However, those who don’t identify highly as white and were made to think about privilege showed less levels of modern racism. The researchers believed this was due to the fact that those people were more susceptible to guilt (which they had explored in a different paper).

Political view also had an impact overall. Those who identified themselves as conservative showed higher levels of modern racism while those who said they were liberal had less modern racism. This was completely separate to whether they spent time thinking about privilege or not.

To me, this study and its findings are fascinating. It reveals a lot about why it is important to research human behaviour. We often think we have clear answers for how people will react to certain situations. Common sense is often wrong, and worse, it can lead to unintended consequences that actually hinder the goal you are trying to achieve.

I’m definitely not saying that thinking about privilege in society should stop. It shouldn’t. It clearly has a role to play. We can see in this study that people who didn’t identify highly as white actually responded well to thinking about their privilege. But it’s the other group of people I’m concerned about. Those are the people who the chant “check your privilege” is so often thrown at. And the last thing we want to do in increase those people’s modern racism, which is so much harder to deal with.

Please let me know your thoughts on this study and topic. I’d be interested in hearing your opinions and whether you’d like to hear more about the large body of research that has gone into changing attitudes and behaviours surrounding racism.


The Main Paper Discussed

Racial attitudes in response to thoughts of white privilege (Nyla Branscombe, Michael Schmitt, and Kristin Schiffhauer)

Some Details on Modern Racism

Modern Racism

Learn More about Privilege

Privilege 101

Social Identity Theory

Simple Psychology – Social Identity Theory



Are Grammar Police really that bad?

Okay, I admit it. I shared the click-bait articles that were floating around a few weeks back about how those people that call out grammar mistakes are actually terrible human beings. It was a moment of weakness. I didn’t read the paper first. I suspected the headline was an exaggeration. But I shared it anyway, in a moment of bitterness over the number of times my own grammatical mis-steps have been nit-picked over. I’m here now to correct that wrong.

This week I’m going to be exploring the paper that sparked all those internet headlines: If You’re House Is Still Available, Send Me an Email: Personality Influences Reactions to Written Errors in Email Messages by Julie Boland and Robin Queen at the University of Michigan Ann Arbor.

There’s a very good reason this paper was so widely shared. A 2011 study out of the University of Pennsylvania reported that aside from a desire to share useful knowledge, there is an emotional element to why certain stories get shared. People love to share stories that trigger a strong emotional response. I suspect that in this case, the strong emotion was vindication. Stay tuned for a closer look at this study in a future Research Ship destination.

A bit of background to the research first. Boland and Queen had previously looked at how the mistakes people make affect how they are judged. Their first paper (I think your going to like me: Exploring the role of errors in email messages on assessments of potential housemates) showed that if you make typos (e.g. teh instead of the) you are going to be judged harshly. But woe betide you if you make grammatical errors (e.g. the dreaded your/you’re slip up). These mistakes were seen as the worst humanly possible when it came to errors of written language.

The results from that study led to another question: Are there differences in the type of people doing the judging? Given that this study was into an entirely new area of research, there wasn’t a lot of background that could guide the research predictions. This makes it tricky to evaluate the findings as there hasn’t been any other research to strengthen their argument. But let’s dig into what they did anyway.

The researchers recruited a reasonable sample of 83 people through Amazon’s crowdsourcing service (Mechanical Turk). Within social psychology it’s fairly standard to have about 20 people in each condition of your study. That means, for every different thing you are testing (in this case, whether they read a passage with no errors, some typos, or some grammatical errors) you should have 20 more people. This study had 3 conditions, so ideally, they should have had around 60. Boland and Queen noted that they used a larger sample size given that there aren’t any similar studies that can be compared with in order to calculate how likely it is that the results you got weren’t just by chance. There’s a lot of really interesting statistical analysis that goes into figuring out how many people you should have in your study, and I’m not going to go into it today, but I wanted to point out that when you see what looks like a small sample size in social psychology it’s usually not that bad.

In terms of how the study was carried out, it was pretty straight forward. People read emails responding to an advertisement for a housemate. The emails either had typos, grammatical errors, or were error free. After reading each email the participants had an opportunity to rate the email writer in several ways. For example, they saw questions like:  “The writer seems friendly”, “The writer seems conscientious”, and “The writer seems trustworthy”. Boland and Queen also had participants fill out a personality survey so they were able to see if there were any particular personality types that jumped out as likely to judge errors more harshly.

The personality scale that was used was the Big Five, which measures personality traits along five different areas: Openness to experience (you like new things), Conscientiousness (you are hardworking), Extraversion (you are outgoing), Agreeableness (you are friendly and avoid confrontation), and Neuroticism (you are easily made anxious). This is a highly respected personality survey with researchers who study individual differences. Boland and Queen didn’t make any overt predictions about what personality traits would be more judgey. But they did say later in the paper, that they had expected those high in conscientiousness to be more aware of typos and grammatical mistakes, though this didn’t come through in the results.

In terms of what the researchers found, there’s a big issue hidden in the statistics, but I’ll save that for after I’ve explained what the was outlined in the paper. If you’ve read the headlines that have been going around the internet you will believe that being a grammar pedant means you are disagreeable and introverted. That’s pretty close to the actual results.

The Findings

Unsurprisingly given Boland and Queen’s previous research, the more typos and grammatical errors that appeared in emails, the less positively people rated the email writer. Both of these findings back up what Boland and Queen found in their last study. So there is something to be said about proof reading.

The researchers did seem shocked that it didn’t matter how old or what level of education the reviewers had when it came to judging typos and grammar mistakes. This seems to have come from an educated guess as I couldn’t find any theory backing up why they would have predicted this.

Weirdly, those who said that they cared a lot about grammatical correctness were more likely to view the email writer more positively when they made typos compared to those who said grammar was less important to them. This seems completely counter-intuitive, but may point to the big issue that I’m going to address shortly.

When Boland and Queen added the personality traits to the analysis, they revealed the findings that the internet has gone crazy for: Those participants who scored high in agreeableness were more likely to judge the email writer more positively when they saw grammatical errors compared to those that scored low in agreeableness. The low agreeableness participants judged the grammar-error filled email more negatively. But there was no difference when the emails had typos in them. The explanation the researchers put forward was that people low in agreeableness may be less tolerant of these mistakes. Although why they would be forgiving of typos is not explained.

Another finding that the media leapt on was how those who scored low in extraversion judged the email. Boland and Queen reported that when those who scored low on extraversion saw typos and grammatical errors, they were more likely judge the email writer more harshly.

There were a couple of other weird findings with conscientiousness, openness and extraversion, but I’m going to stop here to talk about that big issue I mentioned earlier. And that issue is Median Splits. That’s a really jargony term and I’m sorry for using it, but I promise I’m going to explain what’s going on here. Basically, in research you have scales that are continuous – they aren’t grouped in any clear way, and people can slide up and down the scale. These are great scales for getting nuance. Temperature in celsius or fahrenheit is a good example of a continuous scale. Then you have scales that are categorical – these are clear groups of things. People fall into one category, or another. You get a clear distinction, but you lose a lot of nuance. Taking temperature as an example, using the labels cold or hot would be a categorical scale.

The Big Five is typically a continuous scale. People aren’t generally always agreeable or not even remotely agreeable, but rather they fall somewhere along the spectrum in between the two. This is where we get to the big issue. Queen and Boland didn’t treat the Big Five as continuous when they analysed the data. Instead they did what is called a Median Split. They took the very middle point of the personality scale and grouped people either above or below that point. So they artificially created agreeable people and not agreeable people. There are a lot of views about this controversy, but the general consensus amongst statisticians is that it isn’t a good thing to do. Moreover, it makes me very skeptical of the actual results that are reported in this paper.

Let me explain it with the actual agreeableness data that got the internet so worked up. The average score on the agreeableness scale was 3.69 (on a 1-7 scale). This sounds about right, it’s pretty much right in the middle. Now let’s look at what the averages were in the high and low agreeableness groups that Boland and Queen created: The high agreeableness mean was 4.21… that’s only half a point higher than the average overall. The low agreeableness mean was 3.21. The paper didn’t report if these differences were statistically significant, but the fact that they previously used a continuous scale for the demographic data, but decided to split the personality scale up into high and low groups makes me think that their original analysis didn’t pan out how they wanted. So they went back to the drawing board and tried again with a different method until they got the results they ultimately published.


agreeableness scores
Average Low Agreeableness score (orange), Overall Average (black), and High Agreeableness score (green).


I’m more sure that this data was the result of random variance when I see other weird results like extraverts saying that grammar is important to them then rating emails containing typos more positively than extraverts that said grammar wasn’t important to them. Huh?

Now I’m not saying that personality doesn’t play a role in how people assess other people’s language. I’m almost certain that this is an actual effect. But I’m not so certain that this study shows it happening. There definitely needs to be more replication of this study to make sure that we aren’t just seeing a random result made to look like something by massaging the data.

One final note I will leave on is the link the media made between these findings and grammar pedants who point out mistakes. This simply cannot be made from this research, regardless of how sound it is. The participants in this study answered a few questions on what they thought of the email writer, and none were particularly negative or scathing. There were given no opportunity to correct the email writer. And while Boland and Queen did ask participants if they noticed any mistakes in the email, they didn’t report any findings about this question. It’s a huge leap to say that the people who correct your spelling on Facebook are jerks because of this study.

So all that being said. I’m sorry I hastily passed on those articles. Science doesn’t say you are jerk for being part of the Grammar Police. I may still think you are a jerk, I just don’t have the research to back it up… yet. Oh, and I almost certainly made typos and errors in this article, though I did take particular care to proof read. Please point any out to me in the comments and I promise not to call you a jerk.

Did any of you slip up and share this one? Let me know your thoughts on how big this story got.


The Main Paper Discussed

If You’re House Is Still Available, Send Me an Email: Personality Influences Reactions to Written Errors in Email Messages (Julie Boland and Robin Queen)

The First Paper by the Authors

I think your going to like me: Exploring the role of errors in email messages on assessments of potential housemates (Robin Queen and Julie Boland)

Access a version not behind a paywall here

[1] Why people share articles

What makes online Content Viral

The Big Five Personality Traits

A really nice summary of the Big Five from the Personality and Social Dynamics Lab at the University of Oregon 

Issues with Median Splits

A scientific game of telephone

To launch the Research Ship blog I’m going to be looking at Gender on the Brain: A case study of science communication in the new media environment by Cliodhna O’Connor and Helene Joffe from the University College London.

A note about gender – While I acknowledge the spectrum of gender and sex differences that occur, this particular study only reported using CIS male and female participants. I also acknowledge the difference between gender and sex. However, given this week’s research paper is about how science is reported in the media there is a lot of blurring between the language they have used and in order to keep everything clear in terms of paper being discussed I’ll be using the term “gender” when describing the binary sexes, and “male vs female” as the two poles on that binary. 

This week’s paper stood out to me for a couple of reasons. Firstly because I want this blog to be a place where you can read what actually went on in academic research without reading the actual research and I thought a paper showing how the media gets science wrong was a great way to kick off. But mainly it was because I thought it would show me exactly how far off the media are when reporting science and I wanted to gloat. I went into reading it with a pretty heavy bias. While there were times when my opinions were confirmed, for the most part the problems with the scientific findings this article looks at are rooted in the original scientific research, the media only grew on that.

The basic premise of O’Connor and Joffe’s paper was to analyse how the story surrounding a particular published finding was interpreted from its original publication through to press release, traditional media coverage, blog articles, and finally public comments. The methodology of this stood out to me as unique and interesting. While I’ve looked at other text analyses of online behaviour, this was the first time I’d seen researchers following one specific story and mapping how it evolved.  

Let’s start off by setting the stage on the scientific finding they were interested in.

The study O’Connor and Joffe chose to follow was a paper called Sex differences in the structural connectome of the human brain by a group of researchers at Princeton university. A brief outline of their research is that, having noted other biological differences between male and female brains, the researchers wanted to explore any differences that might occur in how these areas connect to one another. They pulled an impressive sample of 949 healthy young adults (428 males, 521 females, aged 8-22) and used a brain imaging technique called diffusion tensor imaging. This technique makes use of the water in our bodies and its tendency to diffuse throughout our tissues, mapping changes in activity throughout the brain. This is a relatively new technique that allows researchers to study the white matter of the brain where as previous imaging techniques were restricted to the grey matter.

Participants were grouped based on their age which allowed the researchers to detect any differences that occurred at different stages of development. These groups were childhood (ages 8-13 1/2 years), adolescence (ages 13 1/2 -17), and early adulthood (ages 18-22). I will point out that the paper didn’t go into any detail on what the participants actually did while they were being scanned. I assume they were told to relax, but I can’t say for sure that any clear direction was given to them while having their brain imaged.

The main findings from the paper were that overall there was a statistically significant difference between the connection patterns in male and female brains. Female brains tended to have connections that cross over each of the hemispheres, while male brains tended to have connections that stayed in one hemisphere or the other. There was also a distinct developmental progressions detected – younger participants had less gender differences. The researchers did point out that the differences were still there, and suggested this was  where different brain development patterns begun to occur.

Overall, the researchers concluded that their study supported previous findings that there were biological differences between male and female brains. They did not make any mention of this confirming stereotypes about either gender. However, they do reference another study carried out on the same group of people that revealed women outperforming men on attention, word and face memory, and social tasks, while the men excelled in motor tasks and had a better understanding of space.

As a final word on their findings the researchers conclude with a discussion about how the difference in brain biology between men and women fits in a complementarity model. That is, that men and women are fundamentally different and this is so they can fit together better in society with each one contributing different innate skills that balances the other out.

Now that we’ve covered the basics of the science that was being communicated, let’s move on to how the communication was actually done.

O’Connor and Joffe were primarily interested in how the findings I just outlined would meander its way through the public consciousness. They began with the original article then moved on to the official press release from Princeton University, then traditional print media (87 articles), blogs discussing the research (162 blogs), and finally user comments on online news websites. The articles and blogs were sourced through pretty standard search databases (including Google). For the readers’ comments section of their analysis the O’Connor and Joffe matched up the print media with their respective websites and collected non-spam comments. Out of the 87 articles, 32 had reader comments available. Given the huge amount of content this turned up, the researchers randomly selected 10% of the comments on each article resulting in 420 individual comments ready to be analysed. I was curious why they did not include comments on blog articles as well, or delve into the wonderful world of social media. So keep in mind when thinking about what was discovered from the readers’ comments that it only included comments on news websites and not Facebook, Twitter, or comments to blog articles, which definitely leads to issues of selection bias

The analysis was broken down into five different areas of interpretation:

  • How does difference in brain connectivity between men and women manifest behaviourally?
  • What is the cause of the difference?
  • How wording was used  to describe the difference between male and female brains?
  • Would this lead to different valuations of men and women?
  • How does this relate to current gender politics?

I’m going to tackle each of these questions separately to keep things nice and clear.

How does difference in brain connectivity between men and women manifest behaviourally?

The original research paper speculated that this biological difference in how male and female brains were connected could be the basis of behavioural differences between genders. O’Connor and Joffe point out that the original paper conducted no behavioural measures so all speculation comes from references other bodies of work that found differences between men and women. These being: sensorimotor skills, spatial navigation, intuition, memory, social cognition, and attention.

The press release carried these ideas forward and mentioned all of the previous areas of study that the original paper speculated would be impacted by the difference in brain connectivity. The press release also introduced the idea that the differences in brain connectivity would explain why men are better at single-tasking and women are better at multi-tasking. The media leapt on this concept and ran with it. This became was the main headline of the findings in the press.

Another point of difference between the press release and the media was that, while most of the areas of domain difference were retained, the speculation that ‘attention’ would be impacted began to decrease. This paper suggests this may be due to attention being mentioned later in the press release and that perhaps the media paid more attention to the earlier paragraphs. The aptness of the trait of attention being dropped because journalists were not paying enough attention is not lost on me.

But the media didn’t stop there. Beyond what was contained in the original scientific paper and the press release the media started to enhance the story with another gender stereotype. The popular dichotomy of emotion and rationality which was linked to the ever-difficult idea of right vs left brain dominance. (I’m definitely going to go into this at some point so stay tuned).

Another diversion from the original article was the inference that this study explained why men and women participate in different forms of labour, specifically domestic chores and childcare responsibility falling primarily towards women while “breadwinning” was a male-wired ability.

By the time we get to the comments the battle of the sexes had completely taken over and the original paper – which I’ll remind you didn’t contain any findings on behavioural differences – is now interpreted as giving proof that men are “wired” to be rational single-taskers who are more productive while women are emotional but are great multi-taskers which makes them idea mothers and housekeepers.

What is the cause of this difference?

As you would expect from good science the original article (and the press release) made no mention of what caused the biological differences in men and women found in the neuroimaging study. It did however spark much debate in the public sphere. The text analysis found that the main angle taken by the media was that it was a biological difference. The metaphor “hard-wired” was used many times suggesting that the differences in biology (and ergo the behaviour they inferred from it) was natural and unchangeable.

However, there was discussion around the topic and not every source was adamant the differences were biological. A number of sources (both traditional news and blogs) discussed potential socialisation differences that may be the cause. Though these numbers were overshadowed by the biological argument.

What is interesting is that this heavy leaning towards biological causality vanished when analysing the public comments. Here, people took equal views on whether the cause of differences in neurobiology was biological or social. And many appreciated the interplay of both, cautioning the journalists for taking a hard line one way or another.

How wording was used  to describe the difference between male and female brains?

O’Connor and Joffe point out that the original paper did not offer any specific explanation as to the ways the sexes were different, but rather as proof that they are different, period. But they are more interested in the way that difference is described. Throughout all the media (including the original scientific paper) the differences were framed as “fundamentally different” using strong adjectives. The analysis didn’t go into why they believed such strong language was used, though I suspect it has to do with getting clicks on an article than any real belief that “men and women might almost be separate species”. Often these vast differences were framed positively. That the difference between men’s brains and women’s brains helped the two to live harmoniously together. But there were examples of this complementary view of the sexes being used by some traditional press and blogs as an example of why homosexuality is not natural.

Would this lead to different valuations of men and women?

As can be expected when a finding is released that reveals a difference between sexes the discussion of ‘which sex is better?’ would arise. As also can be expected the original scientific paper and the press release presented both sexes in a positive light. Also, the large majority of media and blogs did not fall on either side of the gender war. However, there is a clear and disturbing trend when analysing the comments section.

Straight away the analysis revealed that commenters were more likely to favour one sex over the other. This preference was typically towards men. What is more saddening was that when comments were negative, they were overwhelmingly negative towards women.

We should have known by the all too often repeated line: “Never read the comments”.

How does this relate to current gender politics?

A major point O’Connor and Joffe raised was that the research did not show the unbiased view that we expect science to show. While the original scientific paper was not explicit in talking about how their findings upheld traditional gender stereotypes, the media gathered quotes from the researchers that supported this view.

The story being told by this finding was largely a counter to political correctness. Many media used this finding as proof of fundamental differences between men and women. By the time the paper was being blogged about, the finding was being heralded as a way to quiet down pesky feminists. And while a percentage of the news media and blogs made an effort to explain the background of gender discrimination, a sample of the content outright denied this discrimination happened at all.

Overall, the researchers concluded that the main body of media representation heralded the findings as an incredible discovery that validated gender stereotypes. This went on to spark further debate in the blogosphere and comments sections about gender discrimination. And given what O’Connor and Joffe uncovered in comments about the value of men and women you can only imagine how that debate went.

O’Connor and Joffee also make note of the following important points:

The original scientific paper made leaps of logic that were outside of the scope of their study. Particularly when describing the sex differences as ‘fundamental’, when stating that the differences between the sexes made them complementary, and their speculation on the functional outcomes of such differentiation even though they did not study any correlations to behaviour at all.

They also mention that some directions that media went in (gender stereotypes, and parenting ability) were helped by direct quotes from the researchers themselves, again, despite having no real empirical evidence to back this up.

This paper pinpoints the press release as critical for shaping the views the media took when carrying this finding into the public realm. Given that many journalists will only view a press release, it is vital that the important facts of research are included and correct.

That this particular study was interested in gender differences led to a certain type of reading by the media and by the public. This paper mentions previous studies finding that when people are presented with abstract scientific data they are naturally inclined to project stereotypes onto those findings.

There is also previous research into how people can use scientific research to build up differences between groups. A lot of this comes from the ever expanding area of intergroup relationships research and how, as a social species, people feel the need to set up distinctions between what is their group and what is the other group.

I was most taken in this paper by the animosity shown towards women by the commenter but was relieved to read that the overwhelming take on the findings was that the differences in the male and female brains led to them being complementary thus allowing a stronger unit when together. And that both of the areas men and women excel in, though different, was equal.

However, the authors cite other research that suggests viewing gender differences as complementary can work to hide gender inequalities and thus perpetuate discrimination. Given this, it is concerning to me that this is the main story that is being told about this finding – even by the scientists that published it to begin with. Further, while the original paper has issues with its conclusions, the media, by portraying the finding as biological proof gives it a larger amount of weight. Not to mention the how much authority is given to studies that use neuroscientific language and imagery.

All this being said, the authors go on to state that it is not wise to leap from findings in other laboratory controlled experiments and link these to the month long case study they looked at on this one occasion. They point out that despite the heavy biological language used in the original paper and press release, the popular commentary did not latch onto this and brought about further discussion surrounding nature versus nurture. More interestingly, they mention a subset of the data (which they admit was small but not insignificant) that pointed out delving into supposed gender differences may cause more problems in discrimination and were skeptical about the neurobiological arguments the media appears to love.

In terms of this method of research the authors finish with a note about how it would be useful to collect socio-demographic data about the commenters so they can attribute age, gender, location to their comments. I would also like to re-emphasise the need to gather comments on a wider range of platforms, specifically facebook and reddit.

This article got me all fired about about the controversy in gender research and it’s likely I’ll come back to some of the research mentioned in this article. If there are any particular studies you would like me expand upon, please let me know. I’d also be interested in knowing your thoughts about this analysis and the original scientific finding that sparked it. Do you think the media has a responsibility when it comes to reporting gender difference? Do you think scientists have a responsibility when studying it?


The Main Paper Discussed

Gender on the brain: A case study of science communication in the new media environment (Cliodhna O’Connor & Helene Joffe)

The Original Academic Paper

Sex differences in the structural connectome of the human brain ( Ingalhalikar et al.)


Set the sails!

When I decided I wanted to start up my writing again, I tossed around a few different ideas and finally settled on the wonderful world of Science Communication.

The reasons for this are plenty. I am frustrated by the surface level of investigation many science blogs go into. I adore science, learning about science, and telling people about science. But mainly, it’s because it is important.

The halls of academia have long been isolated and shut off to many people in the large unwashed masses. With the rise of the internet, sharing ideas about science and technology has gotten easier and easier, and that’s awesome. But one thing I’ve found was missing from the click-baity headlines, and the 10 Facts You Won’t Believe about Your Brain?! is a  place that takes academic research (both old and new) and talks about it in easy to understand terms that doesn’t take away from the essence of the findings.

So that’s what I’m going to set out to do. Each week I’ll be scouring online and print journals for research papers I find fascinating and discussing them in easy to read language.

I have to do something to keep myself occupied while I wait for grad school acceptance letters.